Kristin Østlie1, MD, PhD, Beate Garfelt1, MD, Ola Skjeldal2, MD, PhD and Per Magnus3, MD, PhD

From the 1Innlandet Hospital Trust, Department of Physical Medicine and Rehabilitation, Ottestad, 2Vestre Viken Hospital Trust, Drammen and 3Norwegian Institute of Public Health, Division of Epidemiology, Oslo, Norway

OBJECTIVE: To increase knowledge of health services utilization and related factors in adult acquired major upper-limb amputees.

DESIGN: Population-based cross-sectional study conducted in Norway.

SUBJECTS: Twohundred and twentyfour adult acquired major upper-limb amputees (57.4% response rate). Mean age was 53.7 years, 83.5% were men and mean post-amputation time was 24.0 years.

METHODS: Data were collected by postal questionnaires. Amputation-related health services utilization, perceived benefit and unmet needs were reported and associations between these factors and background factors were estimated by multivariate logistic regression analyses.

RESULTS: Reported health services utilization was relatively low. The reported benefit of utilized services was generally high. For several non-utilized services, considerable unmet needs for utilization were reported. Several background factors were significantly associated with different aspects of the services examined. A considerable discrepancy between reported pain and utilization of specified pain treatments was noted.

CONCLUSION: The findings of this study suggest that several health services available to acquired major upper-limb amputees should be strengthened. Establishing routines to ensure access to optimal, individualized rehabilitation may prevent disability. We recommend additional, longitudinal studies to further assess the need for mental health care and pain treatment in acquired major upper-limb amputees.

Key words: amputees; upper limb; health services research; patient satisfaction; rehabilitation.

J Rehabil Med 2012; 44: 593–600

Guarantors address: Kristin Østlie, Innlandet Hospital Trust, Department of Physical Medicine and Rehabilitation, Jørgen Jensens vei, NO-2312 Ottestad, Norway. E-mail: kristin.ostlie@sykehuset-innlandet.no

Submitted August 26, 2011; accepted January 31, 2012

Introduction

Major arm amputation causes a loss of function that affects all aspects of life (1, 2), and obtaining functional competence requires a substantial effort by the patient and the therapist(s) during rehabilitation (3, 4). The importance of optimal rehabilitation is emphasized by upper-limb amputees (ULAs) typically being young, healthy individuals who should expect a normal life-span with many years of productive employment and active contribution to society (5–7).

Patient satisfaction is a health outcome often used to evaluate quality of care (4, 8–10). There have been few studies of ULA health services utilization and patient satisfaction, and there is a considerable lack of population-based knowledge. Earlier studies are limited in generalizability due to limited study samples or due to assessing only certain aspects of healthcare. Such studies have been conducted on amputees with phantom pain (4), in samples of veterans and service members (10, 11) and on the satisfaction with prosthetist services (9). Studying service members who sustained major limb amputations during during military operations (n = 158, of which n = 32 ULAs), Pasquina et al. (10) found that army centre inpatient treatment, peer visitors, overall medical care and pain management received particularly high satisfaction ratings. Examining the use of health services and the perceived benefit of such use in Scottish amputees with phantom pain (n = 315, the number of ULAs not specified), Whyte & Carroll (4) found that few amputees made use of available services for general amputation-related problems and that even fewer services were used for pain. The services used were not perceived as helpful.

The associations between patient sociodemographic characteristics and satisfaction with medical care are not well understood (8, 12). A meta-analysis (12) concluded that greater satisfaction was significantly associated with greater age and lower level of education. No overall relationship was found for ethnicity, gender, income or family size. In amputees, there is a paucity of studies addressing this topic. In previous studies, age, level, side and aetiology of amputation have been shown to have little effect on overall satisfaction ratings (9, 10), whereas men and persons with fewer years of education have been shown to be less likely to be satisfied with amputee healthcare (9).

To the author’s knowledge, no population-based study broadly assessing ULAs’ perspectives on available health services has been performed previously. The aims of this population-based study were to describe amputation-related health services utilization in adult acquired major ULAs, to describe the perceived benefit of such use, to estimate the extent of unmet needs for utilization and to estimate associations between demographic and amputation-related factors (background factors) and health services utilization, perceived benefit and reported unmet needs.

Methods

Design and subjects

As part of a larger study (13), we performed a survey with inclusion from 1 October 2006 until 30 May 2008. Eligible amputees from the known population of adult acquired major ULAs in Norway were invited to participate. This amputee population was identified through a combined search of the databases of the 2 companies in Norway that make upper-limb prostheses and of the medical records of 3 major Norwegian hospitals. The inclusion process is described in detail in a previous paper (13). Inclusion criteria were: acquired upper-limb loss, major amputation, adult (age ≥18 years on 1 October 2006), resident in Norway and fluent in spoken and written Norwegian. Exclusion criteria were: physical or mental inability to provide the required information. Demographic and amputation-specific features of the responders (n = 224, 57.4% response) are shown in Table I.

Data collection and variables

Data were collected by postal questionnaires returned by pre-paid post. Each informant received a maximum of two written reminders. The questionnaire was developed by expert clinicians (two physicians, one physiotherapist and one occupational therapist). Prior to data collection, our study was approved by the Norwegian Regional Ethics Committee. Data on demographic features, amputation-related features, pain and utilized health services were collected. Pain was reported as the occurrence of phantom pain, stump pain and musculoskeletal pain during the last 12 months (yes/no) in amputees amputated >12 months before survey. The amputees were asked to indicate which health services they had utilized related to the amputation or related to post-amputation problems from a 19-item list. For utilized services, the amputees were asked to report the perceived benefit on a 5-point Likert scale, where 1 was “completely useless” and 5 was “very useful”. This satisfaction outcome measure is likely to reflect the health service’s perceived effect on the amputee’s symptoms and functional status (8). For non-utilized services, the amputees were asked to report whether they had a wish to utilize such a service (yes/no). Such a wish was interpreted as an indicator of unmet need. Furthermore, the overall satisfaction with all utilized health services was reported on a 5-point Likert scale, where 1 was “very unsatisfied” and 5 was “very satisfied”.

Data analysis

Data were analysed using SPSS version 14.0. 95% confidence intervals (CIs) for percentages were calculated with a binomic formula. To assess the representativeness of our sample, we used independent samples t-test (for continuous variables) and Fisher’s exact test (for frequencies) to compare the questionnaire responders (n = 224) and the questionnaire non-responders (n = 166). Variables compared were: gender, age at survey, place of residence, ethnicity (based on name), unilateral/bilateral amputation, amputation level and prosthetic supply (registered in prosthetist databases). The significance level was set at α = 0.05.

The utilization of each health service (% utilization) and the stated wish to utilize non-utilized services (% wish) were reported. For utilization the number of non-responders was low and non-response was analysed as “no utilization”. The perceived benefit of utilized services was re-coded to a dichotomous variable indicating whether the amputee perceived the service as useful (Likert scale 4–5) or not (Likert scale 1–3) and the recoded perceived benefit was reported (% useful). These analyses were conducted separately in unilateral (n = 214) and bilateral (n = 10) amputees. The occurrence of phantom pain, stump pain and musculoskeletal pain during the last 12 months before survey in amputees amputated > 12 months before survey was reported, and the utilization of specified pain treatments in amputees reporting pain was detailed.

In unilateral amputees, the associations between background factors and: (i) health services utilization; (ii) the perceived benefit of utilized services; and (iii) the reported wish to utilize non-utilized services were estimated using multivariate logistic regression. Such analyses were performed for the 5 most frequently utilized services, the 5 services reported as most useful and the 5 non-utilized services with the highest percentages of reported unmet need, respectively. Background factors analysed were: gender, ethnicity, age at amputation, amputation level, amputation cause, time since the amputation and occupational status at amputation (utilization); gender, ethnicity, age at survey, place of residence at survey, marital status at survey, educational level at survey, occupational status at survey, amputation level, amputation cause, time since the amputation and current prosthesis wear (benefit and need). Furthermore, the pairwise associations between utilization of all 19 health services were analysed. Each association was controlled for the same background factors as for utilization, using multivariate logistic regression. Regression analyses were not performed in bilateral amputees due to the small sample size.

Results

Subjects

A total of 390 eligible amputees were identified. Of these 224 completed questionnaires (57.4% response rate). There were no significant differences between questionnaire responders and non-responders for the variables compared.

Utilization of health services

Tables II and III show that health services utilization was relatively low in major ULAs. The mean number of services utilized was 3.41 in unilateral amputees (95% CI 3.05–3.77, range 0–14) and 3.30 in bilateral amputees (95% CI 1.45–5.15, range 0–8). Among unilateral ULAs, 18.7% (40/214; 95% CI 13.5–23.9) had utilized at least one level of psychologist/psychiatrist treatment and 31.3% (67/214; 95% CI 25.1–37.5) had used at least one type of specified pain treatment. In contrast, no bilateral amputees reported utilization of these treatments. Several statistically significant pairwise associations between utilization of different health services were noted. For instance, ULAs who had utilized psychologist/psychiatrist treatment in in/outpatient care were much more likely to have utilized a community-level psychologist/psychiatrist than ULAs who had not utilized such treatment in in/outpatient care (adjusted odds ratio (aOR) = 20.73) (Table IV).

Overall satisfaction, perceived benefit and unmet need

Fifty-five percent of subjects (95% CI 48.1–61.9) were satisfied with the health services utilized (overall satisfaction), 23.5% were “very satisfied” and 9.0% were “very dissatisfied”. There was no statistically significant difference between unilateral and bilateral amputees. Tables II and III show that the reported benefit of each service was generally high. The most important exceptions include the employment services, the occupational health services, in/outpatient psychologist/psychiatrist treatment and specialized outpatient pain treatment. For several services, especially at the community level, the reported unmet need for utilization was considerable.

Pain and pain treatments

Table V shows that most ULAs reported having experienced at least one type of pain during the last year. Table VI shows that the reported utilization of specified pain treatments in ULAs reporting pain was low. The frequent occurrence of more than one pain type (Table V) did not allow for independent analyses of utilization within each pain type. Excluding treatment non-users, the mean number of pain treatments utilized was 1.4 (95% CI 1.23–1.55). The reported unmet need for utilization of these treatments was low.

Background factors

Significant associations were found between health services utilization and several background factors, e.g. amputation level and aetiology (Table VII). Furthermore, ULAs with ≥ 4 years education and ULAs residing in Western Norway were more likely than those with fewer years of education and those residing in Eastern Norway to report hospital inpatient treatment in a rehabilitation ward as useful (aOR 5.23, 95% CI 1.25–21.96 and aOR 7.55, 95% CI 1.04–54.83, respectively). Current prosthesis wearers were more likely than non-wearers to rate community-level physiotherapy as useful (aOR 4.48, 95% CI 1.08–18.66) and women and ULAs aged ≥ 40 years at survey were more likely than men and ULAs aged 18–39 years to rate community-level occupational therapy as useful (aOR 20.26, 95% CI 1.20–342.27 and aOR 24.17, 95% CI 2.67–218.97, respectively). Moreover, through or above elbow (proximal) ULAs were more likely than transradial ULAs to report a wish to utilize hospital inpatient treatment in a rehabilitation ward (aOR 8.45, 95% CI 1.42–50.31) and community-level occupational therapy (aOR 4.23, 95% CI 1.22–14.69). Non-ethnic Norwegian ULAs were more likely than ethnic Norwegian amputees to report a wish to utilize hospital inpatient treatment in a rehabilitation ward (aOR 17.65, 95% CI 1.23–252.69).

Discussion

Our main findings were that utilization of health services in major ULAs was relatively low, that the utilized services mainly were reported as useful and that, for several non-utilized services, considerable unmet needs for utilization were reported. A considerable discrepancy between reported pain and utilization of specified pain treatments was noted. Several background factors were significantly associated with utilization, reported benefit and reported wish to utilize the health services examined.

Low utilization of health services among amputees was also described by Whyte & Carroll (4), and may, in part, be explained by the amputees lacking information about available services and their potential benefit. The rehabilitation of ULAs in Norway is not standardized (13), and thus access to several services may depend on the amputee’s initiative. Furthermore, healthcare providers may ration access to services such as physiotherapy and occupational therapy, which often depend on referral from a general practitioner (GP). Other factors may include communication barriers (15, 16), waiting lists that are too long, and regional differences in availability of services (13). However, Norway has a well-organized public health system offering health services at no/low cost for the amputee, and issues related to treatment expenses are less likely to explain non-use.

Existing literature suggests that phantom limb pain and residual limb pain diminishes with time, that persistent phantom limb pain is most often intermittent, and that musculoskeletal pain may often be slight, and not require specific pain treatment (15, 17–21). This may, in part, explain the low utilization of specified pain treatments. However, in a recent study (22) present musculoskeletal pain was graded as bothersome, suggesting that for this type of pain, other factors may be more likely to explain non-use. For instance, the amputees may have found other methods to cope with their pain (4, 15, 16). Furthermore, as phantom limb pain frequently is discussed in the literature as intractable, some individuals may give up seeking treatment (4).

In accordance with earlier literature (4), our findings suggest that a GP may provide useful pain treatment. A low satisfaction rating for specialized outpatient pain treatment suggests that this service should be evaluated further. To some extent, this rating may indicate unrealistically high expectations of this treatment (8) in amputees with continued and/or intractable pain, emphasizing the importance of conveying adequate information to the amputees and promoting realistic expectations of the available treatments. The high percentage of amputees rating “other pain treatment” as useful calls for further investigation of the contents of, and indications for, this treatment. It might also be useful to assess the prevalence of amputation neuroma and experiences with neuroma-specific treatment among ULAs reporting stump pain.

Our findings suggest that specialized healthcare in rehabilitation wards, limb centres and private rehabilitation centres is useful in ULA rehabilitation, and that these services should be strengthened. Such strengthening may increase utilization and thus prevent disability (5, 17, 23). Furthermore, the community-level health services were in great demand and should probably be offered to a greater extent. Thus, coordination of different levels and types of health services seems important.

Amputation may have severe consequences in terms of employment (4, 6, 13), and receipt of occupational services may improve the amputees’ prognosis for return to work (17, 23, 24). However, the low reported benefit of the occupational health services and the employment services observed in our study suggest that these services may not have been optimal. It is likely that better cooperation between the specialized rehabilitation services, the employment services and the occupational health services could reduce work-related disability in major ULAs.

The long-term association between acquired upper-limb loss and mental health is not clear (25–27). It is, however, known that amputees frequently manifest emotional reactions such as grief, anxiety and depression (1, 28), and our findings regarding psychologist/psychiatrist utilization suggest that post-amputation mental distress should be recognized in ULA rehabilitation. A strong association between community-level and in/outpatient psychologist/psychiatrist treatment suggest that these two service levels supplement each other. The observed differences regarding utilization and perceived benefit may be explained by community-level mental health care often being offered later and for a longer time period than in/outpatient evaluations conducted shortly post-amputation. Furthermore, the community-level services may mainly be offered to ULAs identified as in need of further services when still in in/outpatient care. A spontaneous decline of symptoms with increasing post-amputation time (1, 28, 29) may lead to a more favourable view of the continued treatment. Factors such as lack of resources (i.e. lack of a community psychologist/psychiatrist) and the use of other mental health services (e.g. a psychiatric nurse) may also have affected our findings. We suggest additional, longitudinal studies to further assess the need for mental health services in acquired major ULAs.

Peer visitation may facilitate the rehabilitation process by the peer demonstrating practical skills and giving advice regarding accomplishment of everyday activities. Earlier studies have found peer visitation to enable patients better to cope with depression, fear and helplessness. A coordinated multidisciplinary approach balancing professional therapy with peer therapy has been demonstrated as the most effective (10, 30). Our findings of a low utilization rate and a considerable unmet need for this service, especially among bilateral ULAs, suggest that peer support should be strengthened in future ULA rehabilitation.

The results of the present study suggest that surgery to ease prosthesis use (i.e. stump shortening/lengthening or osseointegration (16, 31, 32)) may be useful to major ULAs. As expected, only a few amputees had undergone grip-promoting surgery (33). Low utilization of home care and home nursing care may be explained by ULAs often performing everyday activities one-handedly or with prostheses (3, 17).

Our findings regarding gender and educational level are in accordance with earlier literature addressing associations between background factors and satisfaction with healthcare in amputees (9). However, our study also provides new information. We found that the utilization and the reported unmet need for utilization of rehabilitation wards and community-level occupational therapy was higher in proximal than in transradial ULAs, suggesting a need for extra rehabilitation in these ULAs. Furthermore, our findings suggest that more frequent referral to the employment services should be considered in amputees over 40 years of age at amputation, and that community-level occupational therapy might be especially useful in this age group. Also, more frequent referral to rehabilitation wards should be considered among non-ethnic Norwegian amputees. It seems that cause of amputation may affect health services utilization, possibly due to differences in co-morbidity between traumatic and non-traumatic ULAs. The observed negative association between post-amputation time and utilization of several health services may reflect an increasing availability of services during the past decades. We recommend further studies on the importance of these and other background factors to ULA health services utilization and satisfaction.

It is likely that our findings may apply to other western countries (13), with caution regarding differences in healthcare organization. There are, however, some limitations to this study. A relatively low response rate may have produced selection bias; although our analyses of background factors indicate that the questionnaire responders may be considered as representative of the sampling frame, we have no knowledge of health services utilization and experiences with rehabilitation among the non-responders. Information bias may have been introduced by missing data; if those satisfied have a larger chance of responding to questions on benefit, our estimates of service usefulness may be too high. Correspondingly, our estimates of unmet needs for utilization may be too high. Recall bias may have affected reported utilization, and the time between amputation and survey may have affected the reported unmet need for healthcare. Our sample size of bilateral amputees was small and our findings regarding these ULAs should be interpreted with caution. Another probable limitation of this study is the use of a non-standardized satisfaction questionnaire.

In conclusion, our findings suggest that several health services available to acquired major ULAs should be strengthened. Establishing routines to ensure access to optimal rehabilitation tailored to meet individual needs may prevent disability. Such routines should include conveying adequate information to the ULAs about the available services. We recommend additional, longitudinal studies to further assess the need for mental health care and pain treatment in acquired major ULAs.

ACKNOWLEDGEMENTS

This work was funded by Innlandet Hospital Trust and by The Norwegian Extra Foundation for Health and Rehabilitation.

The authors report no conflicts of interest.

References